Amnforge

I hope this helps. A few weeks ago, I put on the alpha cool block for my 3090 ftw3. Followed all the directions for installation. It never booted. my Asus mobo said, load vga bios. Evga said there was something wrong with the card and would rma it for me. I didnt mention that i took the card apart and placed a wb on it (idk if that would have voided the warranty but am thankful they agreed to rma it, i clearly broke the warranty sticker they place over one of the backplate screws) . I did try to take it apart a few times to check the installation, maybe i had shorted it out somehow. Even placed the original fan and heatsink which didnt work. I always use LM but i did use thermal grizzly's shield which is basically red nail polish lol. SO... i know it did not short.

 

After receiving my replacement 3090 ftw3, i also was able to get the ekwb 3080/3090 ftw3 block. I carefully installed it. Didnt miss any pads. It worked without any issues, back plate included in the installation. Since then, I have taken it apart multiple times (I couldnt decide between thermal griz Conductonaut or Kryonaut). Still no issues after 2 weeks. To be honest, the cooling performance is good but under certain circumstances not worth it. In my case, I'm cooling a 5950x right after cooling the 3090. My pc case (TT p8) is kinda limited (15 quad fans and distroplate) when it comes to cooling as i only have space for one TT 360 rad. 

 

When taking into account the undiagnosed issue of the cards that ppl here are experiencing, I wouldnt recommend installing a wb. I dont see the benefits versus the risks at this time. Leave the card as is.

 

I ran my original 3090 ftw3 on fans since I received it in November and think it was pretty sweet looking the way it was.

 

As A true water-head for the last 10 yrs that statement in bold doesn't make sense, in fact it's blasphemous :)
Also, for what its worth I always cut individual thermal pads for each VRAM module.....EK used to always provide those pre-cut squares, now they want us to cut em...not sure whether covering vacant VRAM module locations is a big deal or not, but I always consider the exact way pads are shown to be applied and I would always consider proper and even distribution of the block / pcb in relation to the application instructions....probably no big deal...but..

A last update from me - trying to get the backplate to work, it killed my card. It's now also dead with stock cooler. Rest in peace, 3080 FTW3. See you in RMA heaven.

Filip1995

loztb
A last update from me - trying to get the backplate to work, it killed my card. It's now also dead with stock cooler. Rest in peace, 3080 FTW3. See you in RMA heaven.

Please also reach out to EK about this to investigate your case. Right now the issue doesn’t seem to be caused by wb or backplate. EK has confirmed that they have been testing returned blocks and backplates from effected users on their ftw3 samples. And they could not reproduce the issue some of us are experiencing.

You can be sure EK will get the block and backplate back. I'm not putting anything from EK on a new card.

^Just re-install stock cooler...test it...if still dead RMA

Ok i whil do so xd warming up the card then remove the Stock cooler

And yes i make Some pictures:) i want to cool the card whit two Mora 420 lol

totoro74
Here is another one. Installed block in late Feb. Card worked like dream. Never exceeded 50C in my custom loop. Friday last week I get dreaded 97 Load VGA BIOS code. Long story short I RMAd card today.

Man, how this can be? Did you do anything with your pc prior the issue occuring, like moveing it or installing something in it? Something must have changed for gpu to break down like that.

I installed the EK quantum block with back plate on my 3090 FTW ultra a week ago. Everything went well and booted right up. Hopefully I won't have any future problems.

loztb
Update to my situation - and possible solution for some of you:

After VGA error I installed the EVGA air-cooler and backplate again, in order to RMA the card, but tried it in my computer first - and it booted without any problems.

Ok, so now I know the problem is with my EKWB product, and not with the EVGA card.

I disassemble again, and fit the EKWB block and backplate for another try. Let me assure you, all was put together in perfection, just like the first time.

So, the card gives a VGA error again. Failed POST.

Suspecting something somewhere is shorting the card, I remove the backplate and try again. Still VGA error.

I remove the waterblock once again, and randomly try to eliminate whatever could possibly be causing a short. I cover both rows of capacitors (those next to the VRAM) with a 1mm thermal pad to provide some insulation, and put the card back together.

Guess what, now it boots without any problems.

It could have been a one-off problem with my waterblock, but hopefully this can rescue a few more cards out there. A strip of cheap 1mm thermal pad is a small investment.

This is very interesting!!! Did you mean that you covered the capacitors (w 1mm pad) that are between the Voltage Regulators and the VRAM modules ? or the Capacitors on the other side of the Voltgage Regulators ?

rjbarker

loztb
Update to my situation - and possible solution for some of you:

After VGA error I installed the EVGA air-cooler and backplate again, in order to RMA the card, but tried it in my computer first - and it booted without any problems.

Ok, so now I know the problem is with my EKWB product, and not with the EVGA card.

I disassemble again, and fit the EKWB block and backplate for another try. Let me assure you, all was put together in perfection, just like the first time.

So, the card gives a VGA error again. Failed POST.

Suspecting something somewhere is shorting the card, I remove the backplate and try again. Still VGA error.

I remove the waterblock once again, and randomly try to eliminate whatever could possibly be causing a short. I cover both rows of capacitors (those next to the VRAM) with a 1mm thermal pad to provide some insulation, and put the card back together.

Guess what, now it boots without any problems.

It could have been a one-off problem with my waterblock, but hopefully this can rescue a few more cards out there. A strip of cheap 1mm thermal pad is a small investment.

This is very interesting!!! Did you mean that you covered the capacitors (w 1mm pad) that are between the Voltage Regulators and the VRAM modules ? or the Capacitors on the other side of the Voltgage Regulators ?

 
The original poster later had issues and then reported his card dead.  To me, it seems more like the capacitor was damaged somehow and the addition of the thermal pad closed the break/circuit to allow it to reboot.  He shared a couple of posts after he was able to get this card back up that it failed after continued testing.  Maybe a capacitor/manufacturing issue is more the issue.  

Take it apart, the GPU. Put it back together, with everything stock. Test GPU. Take it apart, put front block on, test GPU. If working then put back block on, then finally test GPU. If all working you're golden.
If anywhere during that process if fails to boot, RMA.
Be careful prying it apart from stock cooler.
Other than that for disassembly instructions look to GN for instruction to take it apart.

I plan on doing that to my second RMA card and documenting it.

GlaucomaPredator
Take it apart, the GPU. Put it back together, with everything stock. Test GPU. Take it apart, put front block on, test GPU. If working then put back block on, then finally test GPU. If all working you're golden.
If anywhere during that process if fails to boot, RMA.
Be careful prying it apart from stock cooler.
Other than that for disassembly instructions look to GN for instruction to take it apart.

I plan on doing that to my second RMA card and documenting it.

Great procedure, has anyone done this yet on a known working card yet? Part of me want to dive in, another wants to wait for someone else to try first. Looking around the rma waiting period can be anywhere from a few days to over 2 weeks, would hate to be down for that long, no replacement gpu.

*Question: Is it okay to run the card with a dry block to test functionality? I assume that much copper can soak up quite a bit of heat, enough to get into the bios or even boot os.

I will not do it again till the cause is found. I used hairdryer on low settings to blow the air between card and the cooler. The whole thing just come apart. Didn't force anything. The card worked for a few weeks and one day just got 97. Thinking back I did tilt the case to get rid of the bubbles from the system but PC was working few days after that. Beside of that PC wasn't moved. Then the day come and it just died. 
No leaks, no worning signs, no overclocking just run stock. The card purely used for gaming at 3440x1440 which would be a child play for that card. I have no explanation why it failed.

So I have an interesting case to add to this thread,
 
Installed my Vector Block on my EVGA FTW3 Ultra 3090 with black lips last week.  Card runs fine until I try to game, at which point the GPU crashes when the card reaches 78 degrees Celsius.  The integrated graphics from the 10900k stay up, and it appears the card has dropped off the hardware set communicating with the MOBO - doesn't appear on device manager.
 
I heated my card with a heat gun prior to pulling off the air cooler, it took awhile and I singed my fingers a bit but ultimately it came apart cleanly with no pieces braking off.
 
For all other purposes the card is working fine, I am typing this right now while it powers my 4 screen setup.  Gpu is reportedly stable running at 32 degrees, but as I stated earlier it crashes once I try to get into a game of Apex or OW (which it ran fine with the stock cooler).  It makes me think that some component is overheating past the max on the card and the card is therefore shutting down on its own accord.  If I reboot the system the card posts and I am able to get to the desktop.  Only pushing the temps on the card up causes crash (given gaming is the point of my 3090 this is still a problem).  I've ordered a bunch of .5mm Thermal pads, as well as some 2mm pads, going to open up the block this weekend and try and see if I can find the piece that is overheating.  I've also ordered a new riser cable since apparently the Thermaltake one thats provided with the P5 is known to cause issues.
 
Build: 
 
Any suggestions on how to find the piece that isn't fitting well?  When the card came from EKWB its not like they had the thermal pads labeled by size, so I spent a bunch of time measuring them against eachother hoping I got the right fit.  I'm praying with actually labeled width pads I can go and redo the thermal padding.
 
Has anyone had their card work with the waterblock only to crash at higher temps like me?  Should I be considering an RMA?  This was my first watercooled build and I would hate to have to go back to the stock cooler 247.

I also noticed discolouration after just few weeks. That on the clear mayhem coolant. Very disappointed. Got an email from EK and definitely not they fault. Even on this forum there are people that done blocks for years and had no issues. I'm pretty sure that there is a fault with block, backplate or pads but they will not admit it ( personal opinion). I will not be buying EK again.